adjust_matrix: Remove rows with low variance and impute missing values

In cola: A Framework for Consensus Partitioning

Description

Remove rows with low variance and impute missing values

Usage

adjust_matrix(m, sd_quantile = 0.05, max_na = 0.25)

Arguments

m	A numeric matrix.
sd_quantile	Cutoff of the quantile of standard deviation. Rows with standard deviation less than it are removed.
max_na	Maximum NA fraction in each row. Rows with NA fraction larger than it are removed.

Details

The function uses impute.knn to impute missing values, then uses adjust_outlier to adjust outliers and removes rows with low standard deviations.

Value

A numeric matrix.

Author(s)

Zuguang Gu <[e-mail]>

Examples

set.seed(123)
m = matrix(rnorm(100), nrow = 10)
m[sample(length(m), 5)] = NA
m[1, ] = 0
m
m2 = adjust_matrix(m)
m2

cola documentation built on Nov. 8, 2020, 8:12 p.m.